To the editor—The welcome criticism of the overwhelming weight given to the 'impact factor', in the December editorial(Nature Neuroscience 1, 641–642) has listed some drawbacks and misuses of this bibliometric method that raise serious concerns. Another adverse effect not mentioned in this editorial is also worth considering. The intense desire to get one's papers published in journals with the highest possible impact factor produces an avalanche of papers submitted to these journals and a progressive disaffection for the less–quoted journals, a self–amplifying process that quickly leads to a diminished reputation and possible financial troubles for the latter. This has the unintended effect of conferring on the editors of highly ranked journals an enormous power of discrimination, no longer regulated by genuine scientific considerations. Increasing submission rates require larger panels of would–be referees, but most editors rely on a list of people they know, who are established experts, hopefully honest and able to provide their reports within reasonable time (a very important point from the editor's point of view). In practice, papers are examined by a rather small number of people, a situation that tends to favor conformity and to result in the suppression of unorthodox results or ideas. Moreover, stringent space limitations imply high rejection rates, and this generates questionable criteria for acceptance; for instance, some editorial offices are instructed to reject a manuscript if the two referees' reports disagree, no matter what the scientific reasons for the disagreement. In other words, there is no longer space for debate, controversial results or divergent opinions. I have experienced this unscientific situation both as author and as referee. Unexpected observations and new ways of thinking vitally need to be submitted to the scientific community for experimental confirmation and discussions. The present concentration of manuscripts in a small number of highly ranked journals is working against these confrontations, so essential for progress in science.